Which one is faster, OpenTSDB or KairosDB?
OpenTSDB is very fast. KairosDB is known as a rewrite of OpenTSDB and is claimed to be even faster than OpenTSDB ( see here ). However, I did some pseudo-distributed cluster tests (1 master, 1 slave, local) for OpenTSDB and 1-node for KairosDB on my VirtualBox (5 GB RAM, 3 cores). The input speed was about 100,000 records / sec for OpenTSDB and 30,000 records / sec for KairosDB. Am I setting up something wrong with KairosDB or is OpenTSDB actually faster?
source to share
I have no measurements on OpenTSDB. We use KairosDB and are pretty fast.
What database did you use for KairosDB? H2 is just for testing and is slowing down desperately. What interface did you use for pushiong data? And if you are using the REST API, how did you structure your requests?
In one bare metal node, we were above 50,000 fps using Telnet (limited by the purchasing agent) and about 3x that rate for gzipped JSON using the REST API (this is for batch inserts of historical data, and JSON is built to insert data with one array of data points for each series and up to 10,000,000 samples per document).
VirtualBox may be slowing down too much (guest VMs have very poor performance).
On the other hand, the last time I spoke to OpenTSDB they were well below 100,000 dpi of insertion ... So they might have better performance.
source to share